Adaptive Frequency Selection for Multi-Band 5G Systems Using Intelligent Algorithms to Maximize Data Rates and Minimize Latency

Abstract

The rapid growth of fifth-generation (5G) networks has introduced complex demands for higher data rates, lower latency, and improved spectrum efficiency. These challenges require intelligent systems capable of dynamically selecting frequency bands across diverse environments. This study presents an adaptive frequency selection framework for multi-band 5G systems that leverages machine learning algorithms to analyze real-time network conditions, user mobility, interference levels, and traffic density. The proposed system operates across sub-6 GHz, Frequency Range 1 (FR1), and Frequency Range 2 (FR2), enabling optimal spectrum utilization based on current operating conditions.

Using the NS-3 network simulator, the framework was evaluated under urban and suburban scenarios. The system achieved spectral efficiencies of up to 15.2 bps/Hz in dense urban areas and 12.8 bps/Hz in suburban settings. Latency was reduced to 15.3 ms in high-mobility environments and 8.2 ms in low-mobility scenarios. Energy consumption was also optimized, showing up to 20% savings compared to static approaches. Additionally, the system improved data rate consistency by 25% and connectivity reliability by 10%. These results highlight the effectiveness of real-time, Artificial intelligence (AI)-driven frequency selection in enhancing 5G network performance under dynamic and complex conditions.